Models of Maternity Care: Updated Evidence on Outcomes and Safety of Planned Home Birth

Women’s and Newborns’ Health Network

February 2011

Copyright to this material produced by the Western Australian Department of Health belongs to the State of Western Australia, under the provisions of the Copyright Act 1968 (C’wth Australia). Apart from any fair dealing for personal, academic, research or non-commercial use, no part may be reproduced without written permission of the Health Networks Branch, Western Australian Department of Health. The Department of Health is under no obligation to grant this permission. Please acknowledge the WA Department of Health when reproducing or quoting material from this source.

Suggested Citation

Department of Health, Western Australia. Models of Maternity Care: Updated Evidence on Outcomes and Safety of Planned Home Birth. Perth: Health Networks Branch, Department of Health, Western Australia; 2011.

Important Disclaimer:

All information and content in this Material is provided in good faith by the WA Department of Health, and is based on sources believed to be reliable and accurate at the time of development. The State of Western Australia, the WA Department of Health and their respective officers, employees and agents, do not accept legal liability or responsibility for the Material, or any consequences arising from its use.

Table of Contents

Authors iv Acknowledgements ...... vii Glossary viii Abbreviations...... xvi Executive Summary ...... 1 1 Review Aims ...... 2 1.2 Assessment of methodological quality of included studies...... 5 1.2.1 Level of evidence...... 6 1.2.2 Assessment of study quality ...... 7 1.2.3 Assessment of statistical conduct...... 7 2 Summary of Evidence Evaluating Planned Home Birth...... 8 2.1 Description of the home birth model of maternity care ...... 8 2.2 Outcomes compared between planned home and hospital birth...... 8 2.3 Maternal characteristics of women planning home birth...... 11 2.4 Maternal satisfaction ...... 11 2.5 Transfers of care ...... 12 2.6 Obstetric interventions and adverse outcomes during labour and birth ...... 12 2.7 Neonatal outcomes ...... 13 2.8 Perinatal mortality...... 14 2.9 Limitations of the evidence...... 15 2.10 Evidence in context of WA Health ...... 16 Conclusions...... 17 References...... 51

Index of Tables

Table 1. NHMRC Levels of Evidence...... 6 Table 2. Levels of evidence and obstetric risk assessed in the reviewed studies...... 10 Table 3. Characteristics of maternity care model in the reviewed studies...... 18 Table 4. Maternal characteristics of women planning home birth...... 23 Table 5. Maternal satisfaction reported by women planning home birth...... 26 Table 6. Antenatal, intrapartum and postpartum transfers from planned home birth into obstetric care...... 27 Table 7. Interventions and outcomes during labour and birth...... 29 Table 8. Neonatal outcomes...... 31 Table 9. Perinatal mortality...... 33

iii

Authors

This evidence-based review was conducted at the Women and Infants Research Foundation, King Edward Memorial Hospital for Women, by the following rewievers:

Dorota A Doherty, BSc (Hons) PhD (Medical Statistics) Dorota Doherty is a Senior Biostatistican and Head of the Biostatistics and Research Design Unit at the Women and Infants Research Foundation. She is also Adjunct Associate Professor at the School of Women’s and Infants’ Health, UWA, located at the King Edward Memorial Hospital campus. Since August 2001, she has provided biostatistical collaboration on research studies conducted at King Edward Memorial Hospital and affiliated institutions. Main area of her expertise is application of biostatistical techniques to the design and analysis of research studies in clinical medicine and public health with particular interest in the area of obstetrics and neonatology. As a biostatistical collaborator, she co-authored many research publications generating evidence in maternity care. Janet Hornbuckle MB ChB MRCOG FRANZCOG Janet Hornbuckle is a Maternal Fetal Medicine Specialist and Clinical Senior Lecturer in Obstetrics and Gynaecology (UWA) working at King Edward Memorial Hospital, Subiaco. Her current clinical responsibilities include management of high-risk obstetric patients in the antenatal and postnatal period within the Maternal Fetal Medicine service as well as regular sessional commitments to the Labour and Birth Suite. She undertakes regular ultrasound sessions in prenatal diagnosis and is the lead obstetrician with responsibility for women with ‘Diabetes in Pregnancy’. Although only involved in tertiary level services she participates in the MSOAP programme which visits rural and remote areas of Western Australia providing education to health practitioners involved in delivering maternity care. Janet relocated from the UK to Western Australia in August 2004 and therefore has experience in other models of Maternity Care. She has considerable experience in reviewing clinical research including publications that specifically involve evidence based reviews: a chapter for the ‘Use of Doppler Ultrasound in Pre-Eclampsia’, a publication ‘Bayesian Interpretation of trials: the example of intrapartum electronic fetal monitoring’ published in BJOG in 2000, and ‘Towards Evidenced Based Management’ in: Intrauterine Growth Restriction: Aetiology and Management, Spinger-Verlag, London 2000. Janet is a regular referee for the BJOG: An International Journal of Obstetrics and Gynaecology and The Australian and New Zealand Journal of Obstetrics and Gynaecology.

Elizabeth Nathan, RN, BSc, PostGrad Dip Biostatistics Elizabeth Nathan is a Biostatistician and works as part of a multidisciplinary team at the Biostatistics and Research Design Unit for the Women and Infants Research Foundation. She was previously qualified as a general nurse. Since 2005, she has been assisting in research relating to the health issues that affect newborns, reproduction and women at all ages through King Edward Memorial Hospital and the School of Women’s and Infants’ Health at the University of Western Australia. She provided statistical support and analysis for the 12th Report of the Perinatal and Infant Mortality Committee of Western Australia published in 2007. She has assisted in the design, conduct and analysis of many research studies and has co-authored fifteen published papers in the area of women’s and infants’ health in collaboration with other investigators. Jennifer J Henderson BSc, PostGrad Dip Adv Nurs (Maternal and Child Health), MPH, PhD Jennifer Henderson is the Project Manager of Preterm Birth research in the School of Women’s and Infants’ Health, The University of Western Australia. She has fifteen years research experience at King Edward Memorial Hospital conducting a wide range of midwifery and obstetric studies. Prior to her research experience she had fifteen years working as a clinical nurse and midwife. Jennifer has experience with both quantitative and qualitative research methodologies. She has worked as a research assistant on a large number of clinical trials including an NHMRC funded trial of a debriefing intervention for postnatal women aiming to prevent postnatal depression. She has contributed to several multicentre randomized controlled trials in the fields of obstetrics and neonatology and was a principal investigator on a Healthway-funded trial of a midwifery home-visiting intervention aiming to improve breastfeeding rates. Jennifer is the recipient of several academic awards. She has a Master of Public Health (with distinction) in which her research topic was the “Effect of Postnatal Depression on Breastfeeding Duration”. She completed her PhD in 2007 in which her research topic was ‘The Effects of Antenatal Glucocorticoids on Lactogenesis in Ewes and Women’. She was awarded the Public Health Association Prize (WA branch) for academic excellence in 2000 and received an NHMRC Postgraduate Research Scholarship for her PhD research. Jennifer has thirteen published research papers with several further articles in progress and 22 published abstracts. Currently she is a principal investigator of the Western Australian arm of a large WHO-funded trial investigating the potential effect of genetic variation on preterm birth.

Correspondence:

Dorota Doherty [email protected] Janet Hornbuckle [email protected] Liz Nathan [email protected] Jennifer Henderson [email protected]

Women and Infants Research Foundation King Edward Memorial Hospital for Women PO Box 134 SUBIACO 6904 WA

Acknowledgements

This project was conducted using the resources of the Women and Infants Research Foundation. We would like to thank Tracy Martin, Principal Midwifery Advisor, Nursing and Midwifery Office at Department of Health WA, for her valuable comments on midwifery care described in this document. We would also like to acknowledge James Humphreys for his assistance with document preparation.

vii

Glossary

Adverse event – a non-beneficial outcome measured in a study of an intervention that may or may not have been caused by the intervention. All or none – all or none of a series of people (case series) with the risk factor(s) experience the outcome. For example, no smallpox develops in the absence of the specific virus; and clear proof of the causal link has come from the disappearance of small pox after large scale vaccination. This is a rare situation. Allocation (or assignment to groups in a study) – the way that subjects are assigned to the different groups in a study (e.g. Drug treatment/placebo; usual treatment/no treatment). This may be by a random method (see randomised controlled trial) or a nonrandom method (see pseudorandomised controlled study). Anaesthetist – a person who is medically qualified to deliver anaesthetics. Analgesia – the relief of pain without causing unconsciousness. Antenatal – existing or occurring before birth (also prenatal). Antenatal care – care of women during pregnancy by doctors and midwives in order to predict and detect problems with the mother or the unborn child. Advice is also offered on other matters relevant to pregnancy and birth. Antenatal clinic – a clinic in a maternity unit where care is provided by midwives, obstetricians and other health professionals. Antepartum haemorrhage – bleeding from the birth canal in the second half of pregnancy. Apgar score – system for assessing the physical condition of infants immediately after birth (a maximum of two points awarded for each of five categories: heart- rate, breathing effort, muscle tone, reflexes and colour). Augmentation of labour – a medical (e.g. Intravenous oxytocin) or surgical (amniotomy) intervention in an attempt to increase the strength of uterine contractions. Best practice in maternity care – care that provides for the best possible outcomes for women and babies in terms of clinical safety and effectiveness. It recognizes that different women have different risks in relation to pregnancy and childbirth. Bias – influences on a study that can lead to invalid conclusions about a treatment or intervention. Bias in research can make a treatment look better or worse than it really is. Bias can even make it look as if the treatment works when it actually doesn’t. Bias can occur by chance or as a result of systematic errors in the design and execution of a study. Bias can occur at different stages in the research process, e.g. in the collection, analysis, interpretation, publication or review of research data. Cardiotocography – the electronic monitoring and recording of the fetal heart rate and uterine activity (CTG). Care giver – a health professional providing services for a client or patient.

viii

Case-control study – patients with a certain outcome or disease and an appropriate group of controls without the outcome or disease are selected (usually with careful consideration of appropriate choice of controls, matching, etc) and then information is obtained on whether the subjects have been exposed to the factor under investigation. Case series – a single group of people exposed to the intervention (factor under study). Post-test – only outcomes after the intervention (factor under study) are recorded in the series of people, so no comparisons can be made. Pre-test/post- test – measures on an outcome are taken before and after the intervention is introduced to a series of people and are then compared (also known as a ‘before- and-after study’). Clinical outcome – an outcome for a study that is defined on the basis of the clinical outcome being studied (e.g. fracture in osteoporosis, peptic ulcer healing and relapse rates). Clinically important effect (see also statistically significant effect) – an outcome that improves the clinical outlook for the patient. The recommendations made in clinical outlook for the patient. The recommendations made in clinical practice guidelines should be both highly statistically significant and clinically important. Cochrane collaboration –an international network that aims to prepare, maintain and disseminate high quality systematic reviews based on randomised controlled trials (RCTs) and when RCTs are not available, the best available evidence from other sources. It promotes the use of explicit methods to minimize bias, and rigorous peer review. Cohort study – an observational study that takes a group (cohort) of patients and follows their progress over time in order to measure outcomes such as disease or mortality rates and make comparisons according to the treatments or interventions that patients received. Thus within the study group, subgroups of patients are identified (from information collected about patients) and these groups are compared with respect to outcome, e.g. comparing mortality between one group that received a specific treatment and one group which did not (or between two groups that received different levels of treatment). Cohorts can be assembled in the present and followed into the future (a ‘concurrent’ or ‘prospective’ cohort study) or identified from past records and followed forward from that time up to the present (a ‘historical’ or ‘retrospective’ cohort study). Because patients are not randomly allocated to subgroups, these subgroups may be quite different in their characteristics and some adjustment must be made when analyzing the results to ensure that the comparison between groups is as fair as possible. Comparative study – a study including a comparison or control group. Confidence interval (CI) – a way of expressing certainty about the findings from a study or group of studies, using statistical techniques. A confidence internal describes a range of possible effects (of a treatment or intervention) that is consistent with the results of a study or group of studies. A wide confidence interval indicates a lack of certainty or precision about the true size of the clinical effect and is seen in studies with too few patients.

Where confidence intervals are narrow they indicate more precise estimates of effects and a larger sample of patients studied. It is usual to interpret a ‘95%’ confidence interval as the range of effects within which we are 95% confident that the true effect lies. Confounding – the measure of a treatment effect is distorted because of differences in variables between the treatment and control groups that are also related to the outcome. For example, if the treatment (or new intervention) is trialed in younger patients then it may appear to be more effective than the comparator, not because it is better, but because the younger patients had better outcomes. Continuity of care – care that helps a woman develop a relationship with the same carer, or group of carers, throughout pregnancy, birth and after the birth. All carers share common ways of working and a common philosophy. The aim is to reduce conflicting advice experienced by women and provide the same philosophy of car throughout the period of care. Continuity of care can be provided in different ways and to varying degrees. Continuous electronic fetal monitoring - the electronic monitoring and recording of the fetal heart rate and uterine activity (CTG). Control group – a group of patients recruited into a study that receives no treatment, a treatment of known effect, or a placebo (dummy treatment), in order to provide a comparison for a group receiving an experimental treatment, such as a new drug. Cross-sectional study – the observation of a defined set of people at a single point in time or time period – a snapshot. This type of study contrasts with a longitudinal study, which follows a set of people over a period of time. Delivery – birth of the baby and the afterbirth. Diabetes – a disorder with high blood sugar levels caused by inappropriate levels of the hormone insulin. Effectiveness – the extent to which an intervention produces favourable outcomes under usual or everyday conditions. Efficacy – the extent to which an intervention produces favourable outcomes under ideally controlled conditions such as in a randomised controlled trial. Epidural (anaesthetic or analgesia) – a local anaesthetic injected around the spinal sac causing some numbness in the lower part of the body. It relieves labour pains effectively. Episiotomy – surgical incision into the perineum and vagina to prevent traumatic tearing during childbirth. Evidence – data about the effectiveness of a new treatment or intervention derived from studies comparing it with an appropriate alternative. Preferably the evidence is derived from a good quality randomised controlled trial, but it may not be. Evidence based – the process of systematically finding, appraising and using research findings as the basis for clinical decisions.

Evidence based clinical practice – evidence-based clinical practice involves making decisions about the care of individual patients based on the best research evidence available rather than basing decisions on personal opinions or common practice (which may not always be evidence based). Evidence-based clinical practice therefore involves integrating individual clinical expertise and patient preferences with the best available evidence from research. Exclusion criteria – see Selection criteria. Experimental study – a research study designed to test whether a treatment or intervention has an effect on the course or outcome of a condition or disease, where the conditions of testing are to some extent under the control of the investigator. Controlled clinical trial and randomised controlled trial are examples of experimental studies. External validity – is the degree to which the results of a study can be applied to situations other than those under consideration by the study, for example, for routine clinical practice. Fetal assessment – assessing and monitoring the fetus during pregnancy. Fetal malpresentation – where the presenting part of the fetus (i.e. the part which is entering the birth canal first) is unusual (e.g. bottom, shoulder, face or brow, instead of the top of the head). Fetus – the unborn baby. Fetal – of fetus. General Practitioner (GP) – a doctor who works from a local surgery to provide medical advice and treatment to patients. Gestation (or gestational age) – length of pregnancy Guidelines – systematically developed statements that assist in decision-making about appropriate health care for specific clinical conditions. Heterogeneity – or lack of homogeneity. The term is used in meta-analyses and systematic reviews when the results or estimates of effects of treatment from separate studies seem to be very different, in terms of the size of treatment effects, or even to the extent that some indicate beneficial and others suggest adverse treatment effects. Such results may occur as a result of differences between studies in terms of the patient populations, outcome measures, definition of variables or duration of follow up. High risk – a term used by clinicians to describe women who have a history of problems in a previous pregnancy or have an existing medical condition or have some potential risk of complications that might require speedy or specialist treatment. Historical controls – data from either a previously published series or previously treated patients at an institution that are used for comparison with a prospectively collected group of patients exposed to the technology or intervention of interest at the same institution. Home birth – usually a planned event where the woman decides to give birth at home, with care usually provided by a qualified health professional. Homogeneity – the results of studies included in a systematic review or meta- analysis are similar and there is no evidence of heterogeneity.

Results are usually regarded as homogeneous when differences between studies could reasonably be expected to occur by chance. Hypertension – blood pressure which is higher than normal, also used for a disease which is characterized by high blood pressure. Incidence – the number of new events (new cases of a disease) in a defined population, within a specified period of time. Inclusion criteria – see Selection criteria. Induction of labour – starting labour artificially by using drugs or other methods. Intention to treat (ITT) – an analysis of a clinical trial where participants are analysed according to the group to which they were initially randomly allocated, regardless of whether or not they dropped out, fully complied with the treatment, or crossed over and received the other treatment. By preserving the original groups one can be more confident that they are comparable. Intervention – clinical procedure in pregnancy or labour e.g. induction or labour, delivery of the fetus with forceps or by caesarean section. Intrapartum – during labour. Labour ward – a suite of rooms set aside in a maternity unit for care of women in labour. Level of evidence – a hierarchy of study evidence that indicates the degree to which bias has been eliminated in the study design. Longitudinal study – a study of the same group of people at more than one point in time. (This type of study contrasts with a cross-sectional study, which observes a defined set of people at a single point in time). Low risk – is a term used by clinicians to describe a woman whose history and condition suggests there is little likelihood of complications. Maternal – relates to the mother. Maternal and Fetal Medicine specialist (MFM) - Obstetrician who specialises in the care of women with high risk pregnancy Meta-analysis – results from a collection of independent studies (investigating the same treatment) are pooled, using statistical techniques to synthesise their findings into a single estimate of a treatment effect. Where studies are not compatible, e.g. because of differences in the study populations or in the outcomes measured, it may be inappropriate or even misleading to statistically pool results in this way. See also Systematic review and Heterogeneity. Midwife – a person appropriately educated and registered to practice midwifery and who provides care, advice and assistance during pregnancy, labour and birth, and after the baby is born. Morbidity – being damaged or diseased. Mortality – number or frequency of deaths. Multiparous – having carried more than one pregnancy to a viable stage. Narcotic – an agent that relieves pain; the term is applied especially to the opioids, i.e. natural or synthetic drugs with morphine-like actions.

xii

Neonatal – refers to the first 28 days of life. Neonatal sepsis – poisoning by micro-organisms growing in the baby. Non-randomised experimental trial – the unit of experimentation (e.g. people, a cluster of people) is allocated to either an intervention group or a control group, using a non-random method such as patient or clinician preference/availability) and the outcomes from each group are compared. Nulliparous – having never given birth to a viable infant. Observational study – in research about disease or treatments, this refers to a study in which nature is allowed to take its course. Changes or differences in one characteristic (e.g. whether or not people received a specific treatment or intervention) are studied in relation to changes or differences in others(s) (e.g. whether or not they died), without the intervention of the investigator. There is a greater risk of selection bias than in experimental studies. Obstetrician – a doctor who specialises in the management and care of pregnant women and childbirth. An obstetrician has specialist education, training and experience and is a fellow of the RANZGOG. Obstetricians provide care in secondary, tertiary and private hospitals. Obstetrics – services relating to the management and care of pregnancy and childbirth, for example antenatal appointments, labour, delivery and care after the baby is born. Odds ratio (OR) – ratio of the odds of the outcome in the treatment group to the corresponding odds in the control group. Operative vaginal delivery – delivery of the baby with the help of forceps or ventouse (vacuum extractor). Paediatrics – a branch of medicine dealing with the development, care and diseases of children. Parous – having borne at least one viable offspring (usually more than 24 weeks of gestation). Peer review – review of a study, service or recommendations by those with similar interests and expertise to the people who produced the study findings or recommendations. Peer reviewers can include professional, patient and carer representatives. Perinatal – refers to the period from 20 weeks of pregnancy to 28 days after birth. Perineum – the area between the vagina and the anus. Postnatal (also postpartum) – pertaining to the four weeks after birth. Postpartum haemorrhage – excess bleeding from the birth canal after birth. Power – see Statistical power. Precision – a measure of how close the estimate is to the true value. It is defined as the inverse of the variance of a measurement or estimate. It is related to the P- value (the smaller P-value, the greater the precision). (Also called statistical precision). Preterm labour – labour occurring at less than 37 completed weeks of pregnancy.

xiii

Prolonged Preterm Rupture of Membranes or Preterm Prelabour Rupture of Membranes (PPROM) – bag of waters breaks or leaks well in advance of the due date and before the commencement of labour. Prospective study – a study in which people are entered into the research and then followed up over a period of time with future events recorded as they happen. This contrasts with studies that are retrospective. Protocols – an adaptation of a clinical guideline or a written statement to meet local conditions and constraints, and which have legal connotations. P-value – the probability (obtained from a statistical test) that the null hypothesis (that there is no treatment effect) is incorrectly rejected. The P-value obtained from a statistical test corresponds to the probability of claiming that there is a treatment effect when in fact there is no real effect (see also statistically significant effect). Qualitative research – is used to explore and understand people’s beliefs, experiences, attitudes, behaviour and interactions. It generates non-numerical data, e.g. a patient’s description of their pain rather than a measure of pain. Quality of evidence (see also strength of evidence) – degree to which bias has been prevented through the design and conduct of research from which evidence is derived. Quantitative research – research that generates numerical data or data that can be converted into numbers, for example clinical trials. Randomisation – a process of allocating participants to treatment or control groups within a controlled trial by using a random mechanism, such as coin toss, random number table, or computer-generated random numbers. Study subjects have an equal chance of being allocated to an intervention or control group thus the two groups are comparable. Randomised controlled trial – a study to test a specific treatment in which people are randomly assigned to two (or more) groups: one (the experimental group) receiving the treatment that is being tested, and the other (the comparison or control group) receiving an alternative treatment, a placebo (dummy treatment) or no treatment. The two groups are followed up to compare differences in outcomes to see how effective the experimental treatment was. (Through randomisation, the groups should be similar in all aspects apart from the treatment they receive during the study). Relative risk or risk ratio (RR) – ratio of the proportions in the treatment and control groups with the outcome. This expresses the risk of the outcome in the treatment group relative to that in the control group. RANZCOG – Royal Australian and New Zealand College of Obstetricians and Gynaecologists. Respiratory distress in the newborn – difficulty in breathing within a few hours of birth. Retrospective study – deals with the present and past and does not involve studying future events. This contrasts with studies that are prospective. Sample – a set of individuals or items selected from the study’s target population so that the hypotheses about the population can be tested.

xiv

Selection bias – error due to systematic differences in characteristics between those who are selected for study and those who are not. It invalidates conclusions and generalizations that might otherwise be drawn from such studies. Selection criteria – explicit standards used by guideline development groups to decide which studies should be included and excluded from consideration as potential sources of evidence. Statistical power – the ability of a study to demonstrate an association or causal relationship between two variables, given that an association exists. For example, 80% power in a clinical trial means that the study has a 80% chance of ending up with a P value of less than 5% in a statistical test (i.e. a statistically significant treatment effect) if there really was an important difference (e.g. 10% versus 5% mortality) between treatments. If the statistical power of a study is low, the study results will be questionable (the study might have been too small to detect any differences). By convention, 80% is an acceptable level of power. See also p value. Statistically significant effect (see also clinically important effect) – an outcome for which the difference between the intervention and control groups is statistically significant (i.e. the P-value is less than 0.05). A statistically significant effect is not necessarily clinically important. Stillbirth – a baby born dead after 20 or 22 completed weeks’ gestation. Strength of evidence – for an intervention effect includes the level (type of studies), quality (how well the studies were designed and performed to eliminate bias) and statistical precision (P-value and confidence interval). Systematic review – a review in which evidence from scientific studies has been identified, appraised and synthesised in a methodical way according to predetermined criteria. May or may not include a meta-analysis. Ultrasound – a diagnostic test which is performed by using ultrasonic waves used to examine the interior organs and structures of the mother and fetus. Uterus – womb. Validity – of measurement: An expression of the degree to which a measurement measures what it purports to measure; it includes construct and content validity. Variable – a measurement that can vary within a study, e.g. the age of participants. Variability is present when differences can be seen between different people or within the same person over time, with respect to any characteristic or feature that can be assessed or measured.

Abbreviations

APH – antepartum haemorrhage CI – confidence interval CTG – cardiotocography gm – grams ITT – Intention to treat IPPM – Intrapartum-related Perinatal Mortality LBW – low birth weight MFM – Maternal and Fetal Medicine specialist N or n – Number of participants in a study sample NHMRC – National Health and Medical Research Council NICU – Neonatal Intensive Care Unit OR – Odds Ratio PE – pre-eclampsia PPH – postpartum haemorrhage PPROM – Prolonged Preterm Rupture of Membranes or Preterm Prelabour Rupture of Membranes RCT – Randomised controlled trial RR – Relative risk vs – versus as in comparison of proportion of women with a defined characteristic (a% vs b%)

xvi

Executive Summary

This evidence-based literature review was undertaken to update the evidence on the safety of planned home birth. The project updated and re-evaluated the evidence regarding community-based midwifery-led care with planned home birth assisted by a qualified practitioner, usually a registered midwife. The debate on safety of planned home birth continues in literature, policy and practice across the developed world. Difficulty in the evaluation of safety of home birth primarily relates to the limitations of evidence that exists in the literature. The evidence consists of observational studies of variable quality which are limited by the lack of appropriate comparison groups, non-representative samples of pregnancies, voluntary reporting of outcomes, small sample sizes and differences between the maternity care models. Since the last review (Henderson, Hornbuckle et al. 2007) several higher quality studies on the safety of planned home birth have been published. These studies involved large cohorts of pregnant women who at the onset of labour were assessed to be at low level of obstetric risk according to standard clinical guidelines. This literature review considered the evidence on planned home birth compared with planned hospital birth. The outcomes examined included characteristics of women electing to have a home birth, maternal satisfaction, antenatal referrals to hospital care, intrapartum and postpartum transfers to hospital, interventions in labour, and neonatal outcomes. The most important outcome considered was perinatal mortality, often reported separately as intrapartum mortality, stillbirth and neonatal death. With no evidence of adverse outcomes associated with planned home birth in low risk pregnancy, this updated review shows that planned home birth with a qualified home birth practitioner is a safe alternative for women determined to be at low risk pregnancy complications by established screening criteria. Women should be counselled about the potential for transfer to hospital if complications arise and systems should be put in place for a smooth transition to hospital care in the case of complications. For women who are not determined to be at low risk, particularly at the onset of labour, there appears to be an excess neonatal morbidity and mortality associated with actual home birth.

Level of evidence: III-2 to IV

1 Review Aims

We have previously reviewed the evidence on models of maternity care including homebirth (Henderson, Hornbuckle et al. 2007). The original review was undertaken to assist with evidence-based planning for Improving Maternity Services: Working Together Across Western Australia (Department of Health 2007). This evidence- based literature review was conducted in order to evaluate the evidence for models of maternity care that may be considered applicable to Western Australia. The review evaluated the evidence on community-based midwifery with planned home birth as one of the models of care, and concluded that it is a safe model of care for women at low obstetric risk by established screening criteria. The debate about the safety of home birth continues in literature, professional policy and practice. Difficulty in the assessment of safety of home birth primarily relates to the limitations of evidence that exists in the literature. Robust and unbiased safety comparisons between the planned home births and planned hospital births require representative samples of appropriately selected women and who received appropriate level of care, for example pregnant women who planned for home or hospital birth should be at low risk of pregnancy complications determined by established obstetric management guidelines. Pregnancy care in women who planned home birth should be provided by registered midwife practitioners operating within clearly defined clinical guidelines, with links for referrals and transfers of care at any stage of pregnancy and birth should the pregnancy no longer be considered at low risk. Women with planned hospital birth may receive hospital or community based antenatal care with the expectation that the birth will occur in either a birth centre or hospital setting. Outcome comparisons between planned home birth and planned hospital birth should be performed according to the planned, not actual, birth setting. The failure to exclude any unplanned home births will overestimate the risk of adverse outcomes (Declercq, Paine at al. 1995), while the failure to exclude planned home births that occurred in hospital after intrapartum transfer will underestimate the risk of adverse outcomes in the planned home birth. Prospective studies that compare planned home and hospital births are often based on small samples of pregnancies and too small to detect any differences in rare adverse outcomes such as perinatal mortality, while several reported large sample size studies that used birth registry data are likely to have a limited ability to completely distinguish between planned, unplanned and actual home births. Several new studies have reported comparisons between planned home and planned hospital births since the last review. The current update of the evidence was conducted to re-evaluate the evidence of safety of planned home birth intended to occur at home with the assistance of a qualified practitioner, usually a registered midwife. With the international debate on safety of planned home birth ongoing along with the desire to provide quality evidence on the planned birth setting, more recent studies involve large pregnancy cohorts more precisely assessed for low level of obstetric risk at onset of labour, where care is provided by qualified health practitioners. This update of the literature follows the same format as the original review. The evidence includes reports from developed countries with comparable obstetric populations and outcomes (Canada, Netherlands, Sweden, UK, USA and Australia). The report combines the data from both reviews and covers the time period between

1996 and January 2011. Several research studies published prior to 1996 and recent small descriptive studies of home birth were also reviewed since they report on home births in Australia.

1.1 Literature search strategy

A comprehensive search of the electronic databases MEDLINE, EMBASE and the Cumulative Index of Nursing and Allied Health Literature (CINAHL) was conducted to identify relevant articles published from January 2007 to December 2010. The search was limited to keywords including ‘home childbirth’, ‘home birth’, ‘home delivery’, ‘booked/intended/planned home birth’ and ‘perinatal mortality’. Additional limits to English language publications and human studies were imposed on the keyword search. A total of 57 citations were identified in the initial search. Studies were excluded after review of title and/or abstract if they were not conducted in developed Western countries, did not use a quantitative methodology in order to evaluate maternal and perinatal outcomes associated with home births and were not published in peer- reviewed literature. Reference lists from each reviewed article were also manually scanned for additional relevant research studies. A total of 13 papers were eligible, a further 4 studies were excluded after detailed evaluation, leaving 9 studies for final review in this report. All research studies included in this review were peer-reviewed quantitative comparisons of pregnancy outcomes between planned home and planned hospital births. All studies differentiated between planned and unplanned home births, although several studies also reported their inability to precisely distinguish between planned and unplanned home births as one of the study limitations and possible source for bias. All articles deemed to be relevant were reviewed independently by a minimum of two of the reviewers. In the original review of evidence (Henderson, Hornbuckle et al. 2007), a total of 48 studies identified evaluating home birth were identified, 28 studies were excluded after initial review and 20 studies were retained for final review. The current update includes the review of 9 further studies giving a total of 29 studies that evaluated safety of planned home birth. 1.1.1 Study exclusion criteria In the current literature review 57 research articles on home birth were published between January 2007 and December 2010, and 48 manuscripts identified in the literature search were excluded from the review. These articles were excluded if they were non-English, non-research articles, not published in peer-reviewed journals, available as abstracts only, qualitative assessments of home birth only, or if they were limited to descriptive data on home births and lacking comparisons with hospital births. Studies with major methodological flaws were also excluded from the review. Articles published only in abstract form, such as conference proceedings, were not deemed to be suitable for this review. Programs aimed at reducing maternal or perinatal mortality in developing countries were excluded because they are rare events in Western Australian maternity care facilities and general health care issues are not comparable to our population. Publications that used a qualitative study design were excluded from the review.

While it is acknowledged that qualitative research studies offer valuable insight into the experience of women planning home birth, these studies provide information regarding reasons for planning home birth but not evidence for evaluation of safety of home birth. 1.1.2 Studies included in the review The review update included studies that evaluated the outcomes of planned home births, irrespective of the actual place of birth, and included a comparison group in which the birth was planned in a hospital. Planned home and planned hospital birth groups were as comparable as possible, that is, the majority of women in both groups were at low obstetric risk at booking. Studies that controlled for differences between the comparison groups in the statistical analysis were also included. The recently published articles for review included five retrospective cohort studies, one retrospective case series, and one prospective comparative study. In the original review of evidence (Henderson, Hornbuckle et al. 2007), the majority of studies were retrospective, observational studies with three prospective cohort studies. There was one attempted randomised controlled trial, which failed to recruit sufficient subjects for a conclusive analysis and was abandoned as unfeasible (Dowswell, Thornton et al. 1996). 1.1.3 Studies reviewed but excluded from the review update Research studies that did not distinguish between planned and unplanned home births were excluded from this review. This was a common limitation of large, population-based studies that relied on data from central databases. Intended place of birth as opposed to actual place of birth was often not recorded. The reason for the exclusion of these studies is because unplanned home births, frequently occurring in women who had little or no antenatal care and who therefore were at higher obstetric risk, are known to have higher rates of adverse outcomes (Declercq, Paine et al. 1995). The inclusion of such studies would mask any putative positive outcomes of planned home births or worsen the apparent incidence of any adverse outcomes. Studies excluded because of the inability to distinguish between planned and unplanned home birth include a large US retrospective cohort of 745,690 births (Wax, Pinette at al. 2010), and a Swedish case-control study report (Hildingsson, Lingren et al. 2006). A UK based study from Scotland that evaluated differences between pregnancy outcomes in planned home births among women who chose pregnancy care provided by independently practicing midwives and women who received the National Health Service (NHS) maternity care (Symon, Winter at al. 2009) was excluded after review. In this study, the outcomes for women under NHS maternity care included home and hospital births. Despite the limitation of this study, the clinical outcomes compared are of interest in our review. The outcomes concerning normal birth, spontaneous labour and perineal trauma were significantly better in the home birth pregnancies assisted by independently practising midwives. The perinatal mortality was similar between home birth with the independent midwives and NHS maternity care in low risk pregnancies. However, for planned home birth assisted by independent midwives, perinatal mortality was significantly worse in pregnancies that were at high risk (multiple pregnancy) or became high risk (breech presentation) compared to those considered low risk when cared for in the same model.

The recently published meta-analysis of observational studies comparing planned home birth versus planned hospital birth was excluded after review (Wax, Lucas et al. 2010). Wax et al. (2010) performed a meta-analysis of 12 studies reporting on pregnancy outcomes that occurred between 1976 and 2006. The meta-analysis concluded no differences in perinatal mortality, but significantly increased neonatal mortality for planned home birth. This meta-analysis has several methodological flaws that are particularly important when combining results from observational studies where matching for confounders is not likely to be adequate. In such instances, a detailed evaluation of quality of all studies is essential; this was not sufficiently described in the manuscript. Moreover, not all studies were included in analyses of perinatal mortality as reporting of perinatal mortality differed across studies. Wax et al. (2010) evaluated neonatal deaths using 6 observational studies that reported neonatal deaths until 28 days of age (Woodcock, Read et al. 1994; Ackermann-Liebrich, Voegli et al. 1996; Janssen, Lee et al. 2002; Pang, Heffelfinger et al. 2002; Lindgren, Radestad et al. 2008). The authors also appear to have included another study that reported neonatal mortality (Hutton, Reitsma et al. 2009), but this was not explicitly stated in the review. The neonatal mortality data were available mainly from small studies, and one large retrospective study of birth registry data where unplanned home births may have been misclassified as planned births because birth certificates used in the study may have not distinguish between all planned and unplanned births, and where the qualification of birth attendant was not always known (Pang, Heffelfinger et al. 2002). Other large studies included in the meta-analysis of maternal and neonatal outcomes (de Jonge, van der Goes et al. 2009; Janssen, Saxell et al. 2009) were not included in the evaluation of neonatal mortality. De Jonge et al. (2009) only reported neonatal deaths within the first 24 hours and 7 days after birth, Janssen et al. (2009) only reported perinatal mortality. Both studies present the data on recent planned home birth outcomes within the Canadian and Dutch midwifery-led care for low risk women and provide best evidence for the Australian setting. The publication of the meta-analysis by Wax et al. (2010) was followed with several editorials commenting on risks of home birth (such as the Lancet editorial: ‘Home birth - proceed with caution’, 2010) and critiques of the study limitations and validity of its conclusions (e.g. Gyte, Dowell et al. 2010; Keirse 2010). One of the crucial questions raised is whether meta-analysis is a correct tool for analysis of observational studies that describe planned home birth within very different healthcare systems (Keirse 2010). The present review update excludes the meta-analysis by Wax et al. (2010) given its methodological limitations and questionable external validity; however, it includes 9 of the 12 studies used by Wax et al. (2010) with the exceptions of the US study conducted between 1976 and 1982 (Koehler, Solomon et al. 1984), the UK study conducted between 1976 and 1982 (Shearer 1985) both deemed not sufficiently recent even at the time of our original review (Henderson, Hornbuckle et al. 2007). The failed randomised trial attempted in 1995 in the UK abandoned after recruiting 11 women has also been excluded for this review (Dowswell, Thornton et al. 1996).

1.2 Assessment of methodological quality of included studies Evidence published in peer-reviewed journals that evaluated home births was of mixed quality. There was considerable variability in the quality study designs and outcomes in the identified studies.

Evidence of planned home birth safety is based on observational studies. Randomised controlled trials of planned home birth versus planned hospital birth were shown to be unfeasible because of strong preferences for either a home or hospital birth by potential recruits which precluded the possibility of randomisation (Dowswell, Thornton et al. 1996). Since the last review, an attempt to conduct a randomised controlled failed in the Netherlands as the women were unwilling to be allocated place of birth at random because of strong preference for either home or hospital setting (Hendrix, Van Horck et al. 2009). The initial review of home births included observational studies evaluating pregnancies that occurred until year 2000 and published until 2006. The current review adds research studies comparing outcomes in pregnancies that occurred between 2000 and 2006 and published until 2010. All studies included in this review, non-randomised comparative studies and patient series, were assessed for their methodological quality. Assessment of research studies followed recommendations for developers of guidelines (NHMRC 2007) with modifications appropriate for a literature review. Three dimensions of the evidence strength were assessed:

 level of evidence, reflecting the effectiveness of study design in its ability to answer research question,

 methodological quality evaluating a likelihood of bias influencing results,

 quality of the statistical conduct of the study. 1.2.1 Level of evidence Level of evidence was designated according to the NHMRC guidelines (Table 1). Levels of evidence for interventions were assigned when comparisons of standard maternity care against alternative models of care were made.

Table 1. NHMRC Levels of Evidence Level Intervention I Systematic review of level II studies II Randomised controlled trial III-1 Pseudo-randomised controlled trial III-2 Comparative study with concurrent controls: . Non-randomised experimental trial . Cohort study . Case-control study . Interrupted time series with a control group III-3 Comparative study without concurrent controls: . Historical control group . Two or more single arm study . Interrupted time series without parallel control group IV Case series with either post-test or pre-test/post-test outcomes Based on NHMRC levels of evidence for intervention (NHMRC 2007).

1.2.2 Assessment of study quality Quality of studies was rated as low, medium or high according to the following considerations:

 Patient selection and inclusion/exclusion criteria

 Comparability of the groups (i.e. obstetric risk and demographic profiles)

 Completeness of follow-up

 Any other feature of study designs that may have introduced bias. 1.2.3 Assessment of statistical conduct Rating of the conduct of statistical analysis as low, medium or high was given according the following criteria:

 Number of participants and statistical power for outcomes considered

 Adequacy of the study sample description

 Intention to treat analysis (in randomised controlled trials)

 Inclusion in the analysis of all selected cases (case-control studies)

 Appropriateness of the statistical analysis (i.e. adjustments for group differences when appropriate)

 External validity of the study results (do the results apply to populations other than the study sample?) Several important surveys and government reports on patient outcomes (level of evidence > IV) were included in the review without formal tabulation of quality assessments.

2 Summary of Evidence Evaluating Planned Home Birth

Less than 1% of women in Australia elect to have home births. In a few countries the incidence of home birth is high, for example 30% of all births in the Netherlands and 3% of all births in the UK. In those countries the infrastructure for safe home birthing is well established and outcomes are generally positive. In other countries such as Canada and USA, the incidence of home births continues to rise. The controversy about the safety of births at home is ongoing. This review evaluates the evidence regarding the safety of planned home births.

2.1 Description of the home birth model of maternity care Planned home birth refers to births that are intended to occur at home with the assistance of a qualified practitioner, usually a registered midwife. Generally, women who plan home birth are at low risk of obstetric complications. The definition of low obstetric risk is often defined within established clinical management guidelines and most often includes women without medical complications, such as hypertension or diabetes, and with an uncomplicated pregnancy, gestational age 37 to 41 weeks at onset of labour, with a single fetus with a vertex presentation. In countries where planned home births are endorsed, such as the Netherlands, the independent practising midwives only provide home birth care to women at low risk at onset of labour as defined by the Obstetric Manual of classification of referrals to secondary or tertiary care (Amelink-Verburg, Verloove-Vanhorick et al. 2007). In Canada, the midwifery-led maternity care with an option for planned home or hospital birth is provided by the registered midwives that operate with a single-payer universal health care system within a province (British Columbia: Janssen, Saxell et al. 2009; Ontario: Hutton, Reitsma et al. 2009). For example in British Columbia and Ontario midwifery care is funded by the provincial Ministry of Health and is accessible to all women in the province who meet the eligibility criteria for low obstetric risk defined by the provincial Colleges of Midwives including descriptions of mandatory transfers to obstetric care. In Western Australia, pregnant women can access community-led midwifery care with planned home birth through Community Midwifery WA (CMWA) that operates within the WA Public Health system or with a privately practising midwife (PPM). All practising midwives must be registered with the Australian Health Practitioners Regulatory Authority (AHPRA). PPMs who wish to supervise home births must give notification of their intention to practice to the WA Director of Public Health. CMWA midwives are employees of the WA Department of Health and follow clear guidelines with inclusion and exclusion criteria to ensure that home birth within the publically funded home birth program is only available to women at low risk of obstetric complications. There is no legislation that dictates that only women with low risk of complication may choose home birth with a PPM (referred to as IPMs, Independent Practicing Midwives, in the review of home births; Department of Health WA 2008).

2.2 Outcomes compared between planned home and hospital birth. In this review, planned home births did not include births that were not attended by a qualified practitioner or unplanned births at home. Eligibility criteria for planned home birth in the reviewed studies were generally classed under the heading of low obstetric risk.

The definition of “obstetric risk” varied between studies but most often included women with an uncomplicated pregnancy, gestational age 37 to 41 weeks at onset of labour, with a single fetus with a vertex presentation. If specified, exclusion criteria were usually, but not always, multiple pregnancy, breech presentation, previous caesarean section, medical complications such as hypertension or diabetes, and labour before 37 weeks of pregnancy. In several reviewed studies, no formal inclusion/exclusion criteria for planned home birth were defined and home births did not represent a homogenous group of pregnancies at low risk of complications (for example Kennare, Keirse et al. 2009). Comparison groups in the evaluated studies included women of similar obstetric risk who planned to deliver in hospital. Most studies evaluated only women who remained eligible for home birth at the commencement of labour although some studies also evaluated rates of referrals to hospital care before the onset of labour. Assessment of quality of studies included in this review is shown in Table 2 (page 11). Summaries of models of care compared in each study and definitions of obstetric populations under evaluation are summarised in Table 3 (page 20). This review of evidence on home birth assessed the following pregnancy outcomes: . Characteristics of women electing home birth . Maternal satisfaction . Antenatal referrals to hospital care . Intrapartum and postpartum transfer rates to hospital . Intervention in labour such as analgesia, episiotomy, operative delivery . Neonatal outcomes such as 5-minute Apgar scores, admission to the neonatal nursery . Perinatal mortality, often reported as intrapartum mortality, stillbirth and neonatal death.

Table 2. Levels of evidence and obstetric risk assessed in the reviewed studies. Research publication Total N Study design Obstetric Level Study Analysis Risk Quality Quality Pregnancy outcomes Hutton, Reitsma et al. 2009† 13,384 R cohort L III-2 medium medium

Janssen, Saxell et al. 2009† 12,982 R cohort L III-2 high high Lindgren, Radestad et al. 12,238 R cohort L III-2 medium medium 2008† Anthony, Buitendijk et al. 191,471 R cohort L III-3 high high 2005 Johnson and Daviss 2005 5,418 P case series S IV high high

Janssen, Lee et al. 2002† 2,176 P cohort L III-2 medium medium Pang, Heffelfinger et al. 17,086 R cohort L III-3 medium medium 2002† Murphy and Fullerton 1998 1,404 P case series A IV high medium

Wiegers, van der Zee et al. 1,836 P cohort L III-2 medium medium 1996 Ackermann-Liebrich, Voegli 874 P comparative A III-2 medium low et al. 1996 Woodcock, Read et al. 1994† 3,904 R cohort S III-2 high medium Woodcock, Read et al. 1990 995 R case series S IV medium medium

Crotty, Ramsay et al. 1990 799 R case series A IV medium medium Howe 1988 165 R case series A IV low low

Perinatal mortality de Jonge, van der Goes 529,688 R cohort L III-2 high high 2009 Kennare, Keirse et al. 2009 300,011 R cohort A III-2 high medium Mori, Dougherty et al. 2008 6,314,315 R case series A IV low low Wolleswinkel-van den 342 R comparative L III-3 low low Bosch, Vredevoogd et al. 2002 De Reu, Hijhuis et al. 2000 8,509 R case series A IV high medium Bastian, Keirse et al. 1998 7,002 R case series A IV medium medium

Northern Region Perinatal 2,888 R case series A IV medium medium Mortality Survey 1996 † Outcomes reported include perinatal mortality. R=retrospective; P=prospective. Obstetric Risk: L= low risk defined by the local criteria; A=no definition of risk level; S=singleton pregnancies only, with or without additional criteria for term pregnancy or non-anomalous fetus or breech presentation. Assessment of evidence quality was formally performed for two studies of maternal satisfaction (Wiegers 2009; and Christiaens and Bracke 2009), and one Australian descriptive study of the first 100 women booked for home birth (McMurtrie, Catling-Paul et al. 2009).

2.3 Maternal characteristics of women planning home birth Women who planned home birth tended to have different demographic characteristics compared with women who elected hospital birth. Summary of the statistically significant differences between women who planned home birth and those who planned hospital birth is shown in Table 4 (page 25). Multiparity was the single maternal characteristic reported by all studies that was more frequent in women with planned home birth with the exception of one small Australian study (Crotty, Ramsay et al. 1990). Women who planned for home birth were significantly older in the majority of reviewed studies although maternal age was not recorded in all studies. Other characteristics with statistically significant differences between women who elected home and hospital births included ethnicity, education level, socio-economic status, type of occupation, employment, smoking during pregnancy and BMI. Women who planned home birth were more likely to be Caucasian, have higher socio-economic status, higher level of education attained, currently employed, have normal weight and be non-smokers. In studies from the Netherlands where 30% of women deliver at home (Anthony, Buitendijk et al. 2005; de Jonge, van der Goes 2009), the women who planned home birth were more likely to be of Dutch origin, and were also less likely to be living in large cities (Anthony, Buitendijk et al. 2005). In Sweden, where less than 1% women plan home birth, women who planned home birth were more likely to be born in a European country other than Sweden (Lindgren, Radestad et al. 2008).

2.4 Maternal satisfaction Four studies evaluated maternal satisfaction with planned home compared to planned hospital birth (Table 5, page 27). For women who had uncomplicated births, women who had planned a home birth were found to have a higher level of satisfaction. Furthermore, unanticipated transfer to hospital did not lessen women’s satisfaction with the experience (Wiegers, van der Zee et al. 1998). Another study found high levels of satisfaction in all groups, while women who had planned a home birth were more likely to feel competent, responsible, secure and more able to deal with the labour compared with planned hospital births (Janssen, Carty et al. 2006). A recent study that investigated satisfaction with planned setting for childbirth in the context of two maternity care systems in Belgium and Netherlands, also found that women who planned home birth reported higher levels of satisfaction than women who planned hospital birth (Christiaens and Bracke 2009). In addition, Belgian women reported higher satisfaction with maternity care than Dutch women irrespective of planned birth setting. Evaluation of satisfaction with overall quality of care received in different settings by different care providers during pregnancy and birth in the Netherlands indicated high satisfaction levels reported for all pregnancy and birth care settings (Wiegers 2009). However, the most important factor that increased reported quality of care scores during labour and birth was being assisted by their own midwife and knowing their care provider.

2.5 Transfers of care Most of the studies evaluated outcomes of planned home birth for pregnancies where the onset of labour occurred at home, and therefore very limited information was available on antenatal transfers of care. Summary of transfers reported by the seven reviewed studies is presented in Table 6 (page 28). Three studies evaluated the proportion of women who were referred to hospital care during pregnancy with 7.4%, 10% and 30% of women initially booked for home birth referred out of home care before the commencement of labour (Murphy and Fullerton 1998; Woodcock, Read et al. 1990; McMurtrie, Catling-Paul et al. 2009). Reasons for antenatal referral included preterm labour or preterm prelabour rupture of membranes (PPROM), fetal malpresentation (breech), multiple gestation, antepartum haemorrhage (placenta praevia), and medical problems such as hypertension. Seven studies reported the rates of transfer of women who had commenced labour at home. The proportion of women transferring during labour varied between 1.5% and 13.3%. The most common reasons for transfer during labour were failure to progress, concern about fetal well-being (meconium, fetal heart rate irregularities), and maternal request for analgesia. The proportion of women who transferred to hospital after the birth ranged between 0.7% and 6.7%. The most common reasons for postnatal transfer were postpartum haemorrhage, retained placenta and suturing of perineal lacerations. Neonatal transfers occurred in between 0.06% to 1.4% of home births, with the most common reasons being respiratory problems or evaluation of anomalies. Three studies only reported the proportion of women who elected planned home birth and who successfully birthed at home, ranging between 69.4% and 78.8% (Kennare, Keirse et al. 2009; Janssen, Saxell et al. 2009; Crotty, Ramsay et al. 1990).

2.6 Obstetric interventions and adverse outcomes during labour and birth Summaries of reported statistically significant differences between obstetric interventions in labour are shown in Table 7 (page 30). All studies reported fewer obstetric interventions during labour and birth for women who planned home birth. Compared with women who planned hospital birth, women who planned home birth were less likely to: . require induction (4.3%-5% vs 16%-22.3%); . require analgesia (7.7%-14.1% vs 27.6%-48.7%); . have an assisted vaginal delivery (2.0%-4.4% vs 4.4%-13.8%); . have a caesarean section (2.0%-6.4% vs 13.6%-19%); . have an episiotomy (1.0%-15.8% vs 5.9%-76.5%). Two recent studies also reported a significantly lower requirement for augmentation of labour either by the artificial rupture of membranes (ARM) or using oxytocin (ARM: 19.3%-22.4% vs 28.2%-39.6%; and oxytocin: 5.9%-8.2% vs 13.1%-18.4% for home births and hospital births respectively) (Hutton, Reitsma et al. 2009; Janssen, Saxell et al. 2009).

A significant reduction in blood loss greater than 500 ml from 11.3% to 9.2% was associated with planned home birth (Hutton, Reitsma et al. 2009). Lower rates of 3rd or 4th degree perineal tears, vaginal tears, and sphincter or rectal ruptures were reported in two recent studies (Lindgren, Radestad et al. 2008; Janssen, Saxell et al. 2009).

2.7 Neonatal outcomes Main neonatal outcomes considered in the reviewed studies included Apgar scores, requirement and/or methods of resuscitation, and admission to a neonatal intensive care unit (NICU). Several studies also evaluated the incidence of birthweight <2500 gm (Hutton, Reitsma et al. 2009; Janssen, Lee et al. 2002; Woodcock, Read et al. 1994). The birthweight<2500 gm was not interpreted as resulting of a birth setting, but an indicator of a failure to adequately screen for extremes of birthweight and to plan for appropriate birth setting (Hutton, Reitsma et al. 2009; Janssen, Lee et al. 2002). Reported differences in neonatal outcomes varied between the review studies. Summaries of significant differences in neonatal outcomes reported are shown in Table 8 (page 32). Several studies reported better neonatal outcomes in planned home birth including: . lower rates of 1-minute Apgar scores <7 (Hutton, Reitsma et al. 2009; Janssen, Saxell et al. 2009), . higher mean 5-minute Apgar scores (Ackermann-Liebrich, Voegli et al. 1996), . lower rates of 5-minute Apgar scores<8 (Woodcock, Read et al. 1994), . significantly lower incidence of birth trauma (Janssen, Saxell et al. 2009; Woodcock, Read et al. 1994), . less resuscitation required at birth (Janssen, Saxell et al. 2009; Janssen, Lee et al. 2002; Woodcock, Read et al. 1994), . lower rates of birthweight < 2500 gm (Janssen, Saxell et al. 2009; Janssen, Lee et al. 2002; Woodcock, Read et al. 1994). No differences between the neonatal NICU admissions were found in any of the reviewed studies. All studies except one (Pang, Heffelfinger et al. 2002) found no differences or found favourable neonatal outcomes for planned home birth. Pang et al. (2002) found higher rates of adverse neonatal outcomes in the home birth group (5-minute Apgar scores≤3: 0.4% vs 0.2%, and assisted ventilation for longer than 30 minutes: 0.3% vs 0.2% for home and hospital births respectively). Findings from this study should be viewed with caution because of the high likelihood of false classification of unplanned home births as planned. In this study, births were classified as home births from birth register entries although there was no section for planned place of birth. As the certifier did not have to be a medical or midwifery professional, there was no way of differentiating the type of attendant and some home births may have not been supervised by a qualified birth attendant. Consequently many of the adverse outcomes may have occurred in women who did not plan to deliver at home (detailed critique by Vedam 2003).

2.8 Perinatal mortality Perinatal mortality was assessed overall or was reported as intrapartum mortality, stillbirth and neonatal death (Table 9, page 34). Different mortality indices were reported in different studies. Several studies only reported perinatal mortality either including neonatal deaths in the first 28 days (Janssen, Saxell et al. 2009; Murphy and Fullerton 1998) or including neonatal deaths within the first 24 hours or first 7 days of life (de Jonge, van der Goes et al. 2009). Some studies reported stillbirth and perinatal mortality (Janssen, Lee et al. 2002; Hutton, Reitsma et al. 2009; Lindgren, Radestad et al. 2008) while others concentrated on intrapartum deaths (Kennare, Keirse et al. 2009; Mori, Dougherty et al. 2008). All reviewed studies found no differences in stillbirths (where compared) nor in perinatal mortality between planned home and hospital births. Two studies reported significantly increased risk of intrapartum mortality associated with home birth (Kennare, Keirse et al. 2009; Mori, Dougherty et al. 2008). However, the strength of this evidence must be considered in the context of the limitations in both studies. The increase in intrapartum deaths reported by Mori et al. (2008) was based on the evaluation of intrapartum deaths in pregnancies at varying levels of obstetric risk (detailed discussion reported by Gyte, Dodwell et al. 2009). The investigators were unable to distinguish between unplanned or planned home births, or to adjust for clinical risk factors and therefore the association between planned home birth and increased intrapartum mortality may have been spurious. Kennare et al. (2009) found no difference in perinatal mortality overall (RR=1.38 95%CI 0.56-3.41), and planned home birth was associated with an increased risk of intrapartum death (RR=1.8, 95%CI 1.53-35.87). This increase in was found in a small number of cases only, therefore the estimate has a wide confidence interval limiting the interpretation of the data. The study evaluated planned home births between 1991 and 2006 before the introduction of the Policy for Planned Home Birth in South Australia, and all comparisons were based on pregnancies at all levels of obstetric risk. The investigators reviewed all 9 perinatal deaths that occurred in planned home births. Out of 9 deaths in the planned home birth group, 4 deaths were considered unavoidable regardless of birth setting, 2 deaths occurred in pregnancies where the women were not low risk at booking and 2 deaths occurred in pregnancies where the woman declined intervention when their low risk status changed. Contributing factors such as post-term pregnancy, twin pregnancy, and inadequate surveillance during labour have been identified as responsible for excess mortality in other home birth studies (Bastian, Keirse at al. 1998). Kennare et al. (2009) concluded that adherence to the Policy for Birth at Home would most likely avoid such perinatal deaths. This study is a significant source of Australian data on home births, and it illustrates the importance of ongoing risk assessment and transfers to hospital care. One study found a statistically significant increase in risk of neonatal death associated with home birth (Pang, Heffelfinger et al. 2002). As already discussed in the section on neonatal outcomes (page 14), the findings from this study should be viewed with caution because of the high likelihood of false classification of unplanned home births as planned, and the inability to differentiate the type of birth attendant, including unqualified birth attendants (discussed by Vedam 2003). The possibility that

many of the adverse outcomes may have occurred in women who did not plan to deliver at home needs to be considered while evaluating the strength of this evidence provided by this study. The most recent large studies conducted in the Netherlands (de Jonge, van der Goes et al. 2009) and Canada (Janssen, Saxell et al. 2009; Janssen, Lee et al. 2002) found no excess perinatal mortality to be associated with planned home birth. In these studies, the pregnancy and birth care was offered to the women who plan home birth by practising midwives under the clearly defined management guidelines for screening of pregnancy. All women who planned home birth were at low risk of pregnancy complications at booking and all transfers of care followed standard guidelines.

2.9 Limitations of the evidence Comprehensive comparisons of pregnancy outcomes between planned home birth and hospital birth are subject to many methodological limitations. All evidence is based on observational studies, as the randomisation of women was unfeasible. Uncontrolled studies were limited by a marked potential for selection bias, where women who chose alternative models of care such as home birth or birth centre care were inherently less likely to have pregnancy or birth complications due to their better background health. Women who choose home birth are likely to have different characteristics to those low-risk women who deliver at a hospital, therefore definitive conclusions from any non-randomised studies require analyses that account for these differences in the background characteristics and events during pregnancy, and this is often difficult to achieve in practice. Reviewed studies were conducted in several developed countries where different models of maternity care offered, and where the assessment of obstetric risk varied across studies. In some instances the pregnancies compared were all uniformly at low risk, with transfers of care instigated according to management guidelines, and with both groups, planned home and planned hospital births representing low risk pregnancies, while other studies used surrogate measures for low risk such as singleton term pregnancies. In prospective studies with well selected controls that minimised selection bias, sample sizes were too small to fully assess morbidity and mortality associated with home birth. Larger comparative studies were retrospective, and some of these studies were limited by the inability to comprehensively distinguish between planned and unplanned place of birth or the inability to control for patient characteristics and level of obstetric risk. In studies without mandatory reporting of all births, the compared home and hospital births may have been non-representative of the true obstetric population. The considerable heterogeneity between studies in terms of operating health care models, comparison groups and compared outcomes limits the possibility of conducting meta-analyses of these studies. An important issue to consider is whether the evidence obtained from large sample studies conducted in a health care system that resembles the local circumstances will offer superior evidence over evidence from meta-analysis averaging over dissimilar systems of maternity care. As described in study inclusion/exclusion criteria (page 5), the recently performed meta-analysis that was rejected as a source of conclusive evidence due to the methodological limitations (Wax, Lucas et al. 2010).

The findings of no differences in perinatal mortality must be viewed with caution because fetal or neonatal death is a rare outcome and very large numbers are needed to confidently detect differences between groups. As the evidence regarding planned home births accumulates and the quality of studies improves with more clearly defined levels of risk and use of clear guidelines for consultation and transfer, outcomes including intrapartum, neonatal or perinatal mortality will require re- evaluation.

2.10 Evidence in context of WA Health In the most recent large studies conducted in the Netherlands (de Jonge, van der Goes et al. 2009) and Canada (Janssen, Saxell et al. 2009; Janssen, Lee et al. 2002), pregnancy and birth care were offered to women who plan home birth by qualified, practising midwives under the clearly defined management guidelines for screening in pregnancy. All women who planned home birth were at low risk of pregnancy complications at booking and all transfers of care followed standard guidelines. These studies have been conducted within maternity services similar to that operating within WA Health. The findings of these studies can therefore be generalised to the setting currently in practice in Western Australia. The studies reporting an increased risk of intrapartum mortality and neonatal mortality for planned home birth were not limited to low-risk pregnancies alone. They are therefore not directly applicable to the WA Health publicly funded home birth program where there are clear inclusion and exclusion criteria for women accessing the program along with clear transfer criteria should the woman’s level of obstetric risk change during the pregnancy, labour or birth. WA has a publicly funded model of planned home birth with community-based midwifery care that has clear guidelines of acceptance into the program and transfers for hospital based care where required. The present review shows that this model of care can be considered to be safe for women at low risk of adverse pregnancy outcomes. For this group of women the evidence shows high levels of satisfaction with care, low rate of obstetric interventions such as caesarean delivery, and comparable neonatal outcomes to similar low risk women having a planned hospital birth.

Conclusions

In summary, women who planned a home birth were generally older, better educated, of more affluent socioeconomic backgrounds and consequently may have been at lower risk than women who delivered in hospital. These differences may have influenced the observed beneficial outcomes of home births. On the evidence available, planned home births by women at low obstetric risk were associated with significant reductions in obstetric interventions of labour and delivery, while demonstrating no increases in perinatal morbidity or mortality. Women should be counselled about the potential for transfer to hospital if complications arise and systems should be put in place for smooth transition to hospital care in the case of complications (Davis-Floyd 2003). The evidence shows that for women who are not defined as at low risk, particularly at the onset of labour, there appears to be an excess neonatal morbidity and mortality associated with home birth.

Evidence Based Summary Point Planned home birth with a qualified home birth practitioner is a safe alternative for women determined to be at low obstetric risk by established screening criteria. Women should be counselled about the potential for transfer to hospital if complications arise and systems should be put in place for smooth transition to hospital care in the case of complications. The evidence shows that for women who are not determined at low obstetric risk, particularly at the onset of labour, there appears to be an excess neonatal morbidity and mortality associated with home birth. Level of Evidence: III-2 to IV

Table 3. Characteristics of maternity care model in the reviewed studies. Country Model of care Research Source of data Home birth care Hospital care Inclusions/Exclusions publication provider provider

The Netherlands Health care is free and access to the national health insurance system is universal. Independent primary care midwives provide care to low risk women who may choose to birth at home or in hospital. Home birth care is only provided to women deemed low risk at the onset of labour as defined by the Obstetric Manual of classification of referrals to secondary or tertiary care. If risk factors arise during pregnancy, labour or postpartum, women are referred to an obstetrician for secondary care. Women with normal pregnancies and births are not reimbursed for specialist care. de Jonge, van der Dutch national perinatal Independent primary Independent primary Excludes GA<37 and GA>42 wks, multiple births, pre- Goes 2009 database (M) care midwife care midwife existing medical conditions or obstetric risk factors, prior CS, Obstetrician breech presentation. Also excludes prolonged ROM >24 hrs without contractions, intrauterine death before onset of labour, and congenital abnormality. Anthony, Buitendijk et Dutch national perinatal Independent primary Independent primary Includes all women deemed low risk at commencement of al. 2005 registry (M) care midwife care midwife pregnancy. Obstetrician Excludes multiparous women with obstetric problems in previous pregnancies. Wolleswinkel-van den Medical records of Independent primary Independent primary Includes all women deemed low risk at commencement of Bosch, Vredevoogd perinatal deaths care midwife care midwife pregnancy. et al. 2002 obtained from hospital Obstetrician and midwives’ medical archives (V) De Reu, Hijhuis et al. Records from hospitals, Independent primary Independent primary No exclusions 2000 midwives and GP’s (M) care midwife care midwife General practitioner Physician Wiegers, Keirse et al. Midwifery practice Independent primary Independent primary Includes low risk women only 1996 records (V) care midwife care midwife

Country Model of care Research Source of data Home birth care Hospital care Inclusions/Exclusions publication provider provider

(Source of data: M=mandatory reporting, V=voluntary reporting; Health practitioners described using terminology used in the research publications.) Canada Single payer universal health system. Midwifery care is funded by the provincial Ministry of Health and is accessible to all women who meet the standards for low obstetric risk. Registered midwives provide care for women who choose to home birth and meet eligibility criteria set by the provincial College of Midwives, including descriptions of mandatory transfers to obstetric care. Midwives are well integrated into the health care system, work in individual or group practices, have hospital admission privileges with good access to emergency services, consultation and transfer of care. They provide a model of continuity of care whereby a woman is attended by a small group of midwives throughout pregnancy, birth and the postpartum period. Hutton, Reitsma et al. The Ontario Ministry of Registered midwife Registered midwife Excludes GA<37 or GA>43 wks, multiple birth, pregnancy 2009 Health database of Physician complications, >1 prior CS, breech presentation midwifery care (M) Janssen, Saxell et al. British Columbia Registered midwife Registered midwife Excludes GA<36 and GA>41 wks, multiple birth, pre-existing 2009 Perinatal Database Physician medical conditions, pregnancy complications, >1 prior CS Registry (M) cephalic presentation. Home births include breech presentations determined after onset of labour. Women who had previous CS were not included in planned hospital births. Janssen, Lee et al. British Columbia Registered midwife Registered midwife Excludes GA<37 & GA>41 wks, multiple birth, pre-existing 2002 Perinatal Database Physician medical conditions, pregnancy complications, >1 prior CS, Registry (M) breech presentation

Sweden Home birth is not endorsed by health authorities and is not financed or available within the health care system. Midwives are primary care givers for uncomplicated pregnancies, and care during labour and birth. In complicated cases, midwives cooperate with obstetricians. Lindgren, Radestad Swedish Medical Birth Licensed midwife Obstetrician Planned homebirth group includes pre- and post-term births et al. 2008 Register (M) and multiple births. These were not included in the planned hospital births. Breech presentations were included in both.

19 Switzerland The Swiss health care system is private for all outpatient care. Fees are covered by health insurance, to which everybody subscribes.

Country Model of care Research Source of data Home birth care Hospital care Inclusions/Exclusions publication provider provider

Ackermann-Liebrich, Hospital and home Midwife Midwife No exclusions. The study group had no formal policy on Voegli et al. 1996 delivery records (V) GP Physician accepting women for home birth.

(Source of data: M=mandatory reporting, V=voluntary reporting; Health practitioners described using terminology used in the research publications.) United Kingdom Home birth for low risk women is endorsed by the Royal College of Obstetricians and Gynaecologists and the Royal College of Midwives. NHS community midwives service the majority of women deemed eligible for home birth. Independent midwives are self- employed qualified practitioners working outside the NHS without professional indemnity insurance (since 1994). Mori, Dougherty et Confidential Enquiry No discussion of Not discussed No exclusions al. 2008 into Maternal and Child model of care Includes all births in England and Wales during study period Health (CEMAH) and the Office for National Statistics (M) Northern Region Northern Regional Community midwife Not discussed No exclusions Perinatal Mortality Health Authority Survey 1996 Office of Population Censuses and Surveys (M)

United States Home births are not endorsed by the American College of Obstetricians and Gynecologists. Certified professional midwives attend most home births and out-of-hospital birth centres providing care antenatally, during labour and delivery and post partum. Nurse- midwives are registered nurses who have graduated from a nurse-midwifery education program. They mostly attend hospital births, but also attend home and out-of-hospital births. Johnson and Daviss National Centre for Certified professional Not discussed Excludes GA<37 wks, multiple birth 2005 Health Statistics midwife North American Registry of Midwives (M) 20

Country Model of care Research Source of data Home birth care Hospital care Inclusions/Exclusions publication provider provider

Pang, Heffelfinger et Washington State birth Nurse Not discussed Excludes GA<34 weeks, multiple birth, pre-existing medical al. 2002 registry (M) Midwife conditions, pregnancy complications Physician Murphy and Fullerton Nurse-midwifery Certified nurse- No comparison group No exclusions 1998 practices (V) midwife

(Source of data: M=mandatory reporting, V=voluntary reporting; Health practitioners described using terminology used in the research publications.) Australia Until the late 1990’s, the system in Australia did not provide for planned home births and there were no community midwifery services. Women planning home birth received care from an independent midwife (registered or unregistered) with no formal support from the national health care system or professional indemnity insurance. In South Australia, a publicly funded Community Midwife Project started in 1998. In 2002, Bachelor of Midwifery courses were introduced. Since the introduction of the Policy for Planned Birth at Home in 2007, some hospitals have introduced midwifery group practice models of care with the option for home birth. In Western Australia, the Community Midwifery Program started as a pilot program in 1996, to provide primary midwifery services for home and hospital births. In 2000, the increasing cost of professional indemnity led to the midwives being employed by the Department of Health to allow indemnity through the government insurer. Privately practising midwives continue to practise outside the Community Midwifery Program without professional indemnity insurance for intrapartum care in the home. Since October 2010, professional indemnity insurance is a requirement of registration for all privately practising midwives. This indemnity insurance covers for the provision of antenatal and postnatal care in the community or in a hospital. Kennare, Keirse et al. Pregnancy Outcome Registered midwife Physician No exclusions 2009 Unit of South Australian Health: the collection of perinatal statistics for all births and deaths (M) 21

Country Model of care Research Source of data Home birth care Hospital care Inclusions/Exclusions publication provider provider

Bastian, Keirse et al. Homebirth Australia (V) Independent Physician No exclusions 1998 and national perinatal registered midwife Home births included pre and post-term, multiple births and mortality data (M) General practitioner breech presentations Unregistered midwife Woodcock, Read et Midwives’ Notification Independent Physician Excluded multiple births, congenital anomalies al. 1994 System, Maternal and registered midwife Child Health Research General practitioner Data Base and the Health Department of Western Australia (M) Woodcock, Read et Home birth midwives’ Independent Physician No exclusions al. 1990 records (V), Midwives registered midwife Hospital births include only singleton, Caucasian births Notification System, General practitioner Western Australia (M) Crotty, Ramsay et al. South Australian Independent Physician No exclusions 1990 perinatal data collection registered midwife and medical practice General practitioner case records (V) Howe 1988 Midwives Registry Independent No comparison group No exclusions Western Australia (M) registered midwife

(Source of data: M=mandatory reporting, V=voluntary reporting; Health practitioners described using terminology used in the research publications.)

Table 4. Maternal characteristics of women planning home birth.

Study Years of Country N Results study Statistically significant differences only are reported de Jonge, van der Goes et 2000-2006 Netherlands Home Compared with planned hospital birth, women planning al. 2009 birth:321,303 home birth were more likely to be: Level of evidence III-2 Hospital to be ≥25 years (91% vs 82%), birth:163,261 of Dutch origin (91% vs 65%), multiparous (59% vs 55%), medium-high socio-economic status (81% vs 67%). Kennare, Kierse et al. 2009 1991-2006 South Home birth: 1,141 Compared with planned hospital birth, women planning Australia home birth were more likely to be: Level of evidence III-2 Hospital birth: 297,192 older (31.3 vs 29.2 yrs), higher occupational status, live in metro area (79.8% vs 76.0%); and less likely to be nulliparous (31.2% vs 41.0%) and Indigenous (1.0% vs 2.2%). Janssen, Saxell et al. 2009 2000-2004 British Home birth: 2,889 Compared to planned midwife-led hospital birth women who Columbia, planned home birth were less likely: Level of evidence III-2 Midwife-led Canada hospital birth: to be a single parent (3.1% vs 5.3% vs 3.1%), 4,752 nulliparous (41.9% vs 51.1% vs 41.3%), Physician-led to attend an ultrasound before 20 pregnancy weeks (58.9% hospital birth: vs 70.9% vs 75.5%). 5,331

Christiaens and Bracke 2004-2005 Belgium, Home birth: 230 Compared to Dutch women Belgian women were more likely 2009 Netherlands to have higher education (76% vs 41%) and to be primigravid Hospital birth: 381 (51% vs 41%). Level of evidence III-2 No comparisons between planned home and hospital births. 23

Study Years of Country N Results study Statistically significant differences only are reported Lindgren, Radestad et al. 1992-2004 Sweden Home birth: 897 Women planning home birth were more likely to be: 2008 Hospital >35 years (26% vs 14%), a non-smoker, Level of evidence III-2 birth:11,341 be born in a European country outside Sweden (9% vs 5%), more likely having 2nd or 3rd (57% vs 34%) or having ≥4th child (17% vs 4%), having BMI<25, to be educationally qualified, to be employed; and were less likely to be: <25 years of age (10% vs 19%), non-European born (1% vs 10%). Anthony, Buitendijk et al. 2000 Netherlands Home birth: Multiparous women under the care of a midwife deliver at 2005 160,329 home more often than nulliparous women (43% vs 24%) Level of evidence III-3 Fewer women younger than 25 years Fewer non-Dutch women (17% vs 37%) Fewer women living in large cities (31%) vs regional towns (36%) vs rural areas (36%) Johnson and Daviss 2005 2000 USA and Home birth: 5,418 Compared with all hospital births, women choosing home Canada birth were older, lower socioeconomic status, better Level of evidence IV educated, less likely to be African-American or Hispanic Janssen, Lee et al. 2002 1998-1999 Canada Home birth: 862 Compared with women planning hospital births with a doctor, women planning home births were more often multiparous Level of evidence III-2 Doctor-led hospital (53% vs 42%). birth: 743 Women who planned birth with a midwife were more often Midwife-led multiparous (53% and 52% for planned home and hospital hospital birth: 571 births respectively).

Study Years of Country N Results study Statistically significant differences only are reported Pang, Heffelfinger et al. 1989-1996 USA Home birth: 6,133 Compared with women planning hospital births, women 2002 planning home births were more often: Hospital birth: Level of evidence III-3 10,593 older (≥30 years: 49% vs 33%), white (92% vs 81%), multiparous (76% vs 57%), and less often smokers (10% vs 18%). Woodcock, Read et al. 1981-1987 Western Home birth: 995 Most women choosing home birth were Caucasian and 1990 Australia slightly older. Parity was similar to all Caucasian women having singleton births. Level of evidence IV Crotty, Ramsay et al. 1990 1976-1987 South Home birth: 799 Compared with hospital births, women planning home births Australia were more often: Level of evidence IV older (≥30 years: 38.3% vs 25.1%), with high socioeconomic status (67% vs 38%).

Table 5. Maternal satisfaction reported by women planning home birth.

Study Years of Country N Results study

Wiegers 2009 2007 Netherlands 793 High level of satisfaction with planned home and hospital birth was reported. Level of evidence III-3 Birth assisted by known midwife or knowing care provider increased satisfaction. Christiaens and Bracke 2004-2005 Belgium, 580 Belgian women more satisfied than Dutch women irrespective 2009 Netherlands of birth setting. Level of evidence III-2 Women who planned homebirth were more satisfied than those who planned hospital birth. Janssen, Carty et al. 2006 1998-1999 Canada Home birth: 862 All groups had high level of evidences of satisfaction but women who had planned a home birth more often felt Level of evidence III-2 Physician-led competent, responsible, secure, adequate, relaxed, and more Hospital birth: 743 able to deal with the labour. Midwife-led hospital birth: 571 Wiegers, van der Zee et al. 1990-1992 Netherlands 2,301 Of women who had uncomplicated births, women who had 1998 planned a home birth had a higher level of evidence of satisfaction. Level of evidence III-2 Unanticipated transfer to hospital did not lessen women’s satisfaction with the experience.

Table 6. Antenatal, intrapartum and postpartum transfers from planned home birth into obstetric care. Study Years of Country N Results study Transfer before labour McMurtrie, Catling-Paul et al. 2009 2005-2009 Australia (NSW) 100 30% transferred antenatally

Murphy and Fullerton 1998 1994-1995 USA 1,404 7.4% transferred antenatally Level of evidence IV

Woodcock, Read et al. 1990 1981-1987 Western Australia 995 10% transferred antenatally Level of evidence IV

Transfer to hospital during or after labour McMurtrie, Catling-Paul et al. 2009 2005-2009 Australia (NSW) 100 10.0% intrapartum transfer of care 2.9% maternal transfer after birth 1.4% neonatal transfer Hutton, Reitsma et al. 2009 2003-2006 Ontario, Canada 6,692 78.6% delivered at home without transfers of care Level of evidence III-2 5.4% transferred during or immediately after birth

Johnson, K C and Daviss 2005 2000 USA and Canada 5,418 10.1% transferred in labour Level of evidence IV 1.3% maternal transfer after birth 0.7% neonatal transfer Janssen, Lee et al. 2002 1998-1999 Canada 862 1.5% transferred in labour Level of evidence III-2 0.9% maternal transfer after birth 0.9% neonatal transfer

Murphy and Fullerton 1998 1994-1995 USA 1,404 7.7% transferred in labour Level of evidence IV 0.7% maternal transfer after birth

27 1.1% neonatal transfer

Study Years of Country N Results study Woodcock, Read et al. 1990 1981-1987 Western Australia 995 14.2% transferred in labour or after birth Level of evidence IV

Howe 1988 1983-1986 Western Australia 165 13.3% transferred in labour Level of evidence IV 6.7% transferred after birth 0.06% neonatal transfer No transfers of care*

Kennare, Keirse et al. 2009 1991-2006 South Australia 1,141 69.4% of planned home births occurred at home Level of evidence III-2

Janssen, Saxell et al. 2009 2000-2004 British Columbia, 2,889 78.8% of women who planned home birth delivered Level of evidence III-2 Canada at home.

Crotty, Ramsay et al. 1990 1976-1987 South Australia 799 78.2% of women delivered at home Level of evidence IV 63.8% nulliparous women 87.5% multiparous women * only proportions of women who remained in the planned birth setting were reported.

Table 7. Interventions and outcomes during labour and birth. Study Years of study Country N Results Statistically significant differences only are reported Hutton, Reitsma et al. 2003-2006 Ontario, Canada Planned home birth: Compared with women who delivered in hospital, women who 2009 6,692 planned home birth had lower rates of: Level of evidence III-2 Planned hospital Augmentation: birth: 6,692 ARM (22.4% vs 28.2%) and oxytocin (8.2% vs 13.1%), pharmaceutical pain relief (16.8% vs 54.8%) nitrous oxide (3.3% vs 18.0%), narcotic (1.7% vs 6.3%), epidural (9.8% vs 21.0%); episiotomies (4.3% vs 5.9%), assisted vaginal deliveries (2.9% vs 4.4%), caesarean sections (5.2% vs 8.1%), blood loss>500ml (9.2% vs 11.3%). Janssen, Saxell et al. 2000-2004 British Columbia, Planned home birth: Compared to midwife-led planned hospital birth and physician- 2009 Canada 2,889 led planned hospital birth, women who planned home birth had Level of evidence III-2 Planned midwife-led lower rates of: hospital birth: 4,752 electronic fetal monitoring (13.6% vs 41.9% vs 78.8%), Planned physician- labour augmentation ARM/oxytocin (23.7% vs 39.9% vs led hospital birth: 50.4%), 5,331 ARM (19.3% vs 31.9% vs 39.6%) oxytocin (5.9% vs 12.7% vs 18.4%), assisted vaginal deliveries (3.0% vs 7.2% vs 13.8%), caesarean sections (7.2% vs 10.5% vs 11.0%); nulliparous women (13.0% vs 18.7% vs 21.8%) multiparous (3.0% vs 1.9% vs 3.4%), episiotomies (3.1% vs 6.8% vs 16.9%), 3rd and 4th degree tears (1.2% vs 2.9% vs 3.4%). 29

Study Years of study Country N Results Statistically significant differences only are reported Lindgren, Radestad et al. 1992-2004 Sweden Planned home birth: Compared to women who planned hospital birth, women who 2008 897 planned home birth had fewer: Level of evidence III-2 Planned hospital assisted vaginal deliveries (2.2% vs 9.6%), birth:11,341 cesarean sections (2.4% vs 6.8%), episiotomies (0.8% vs 7.2%), vaginal tears (17.9% vs 31.5%), sphincter/rectal ruptures (0.3% vs 2.7%). Johnson and Daviss 2000 USA and Canada Planned home birth: Compared with hospital births there were lower rates of: 2005 5,418 continuous electronic fetal monitoring (9.6% vs 84.3%), Level of evidence IV episiotomies (2.1% vs 33%), caesarean sections (3.7% vs 19%). Janssen, Lee et al. 2002 1998-1999 Canada Planned home birth: Compared with doctor-attended and midwife-attended hospital Level of evidence III-2 862 births, women who had home births had fewer: Planned doctor-led epidurals (7.7% vs 27.6% vs 26.3%), hospital birth: 743 inductions (4.3% vs 22.3% vs 12.0%), Planned midwife-led caesarean sections (6.4% vs 18.2% vs 11.9%), hospital birth: 571 episiotomies (3.8% vs 15.3% vs 10.9%). Wiegers, Keirse et al. 1990-1993 Netherlands 2,301 Compared with hospital births, multiparous women who had 1996 home births were less likely to have episiotomies (15.8% vs Level of evidence III-2 25.1%) Ackermann-Liebrich, 1989-1991 Switzerland Planned home birth: Compared with planned hospital births, women with planned Voegli et al. 1996 489 home births had less: Level of evidence III-2 Planned hospital inductions (4.6% vs 16.0%), birth: 385 analgesia (17.1% vs 48.7%), instrumental deliveries (4.4 vs 13%), caesarean sections (5.2% vs 13.6%), episiotomies (26% vs 76.5%). 30

Table 8. Neonatal outcomes. Study Years of study Country N Results Statistically significant differences only are reported Hutton, Reitsma et al. 2003-2006 Ontario, Canada Planned home birth: No differences in neonatal outcomes: 5-minute Apgar<7, 2009 6,692 requirement for positive pressure ventilation, NICU>4 days, Level of evidence III-2 Planned hospital birth: birthweight<2500g. 6,692 Janssen, Saxell et al. 2000-2004 British Columbia, Planned home birth: Compared to midwife-led planned hospital birth and 2009 Canada 2,889 physician-led planned hospital birth, women who planned Level of evidence III-2 Planned midwife-led home birth had lower rates of: hospital birth: 4,752 1-minute Apgar score <7 (RR=0.76 95%CI 0.66-0.88 and Planned physician-led RR=0.74 95%CI 0.64-0.86), hospital birth: 5,331 birth trauma* (RR 0.26 95%CI 0.11-0.58 and RR=0.33 95%CI 0.15-0.74), resuscitation at birth** (RR=0.23 95%CI 0.14-0.37 and RR=0.56 95%CI 0.32-0.96), oxygen therapy>24hrs (RR 0.37 95%CI 0.24-0.59 and RR=0.38 95%CI 0.24-0.61). Johnson and Daviss 2000 USA and Canada Planned home birth: 1.3% 5-minute Apgar scores < 7 2005 5,418 2.4% babies admitted to the NICU Level of evidence IV Janssen, Lee et al. 2002 1998-1999 Canada Planned home birth: Compared with physician-led planned hospital birth, women Level of evidence III-2 862 who planned home birth had lower rates of: Planned doctor-led 1-minute Apgar scores<7 (10.4% vs 14.5%), hospital birth: 743 tracheal suction (2.8% vs 5.7%), Planned midwife-led IPPV*** by mask (5.1% vs 8.5%), hospital birth: 571 drugs for resuscitation (0.5% vs 2.7%), birthweight <2,500g (0.8% vs 2.0%), Compared with midwife-led planned hospital birth, women who planned home birth had fewer tracheal suctions (2.8% vs 7.1%). 31

Study Years of study Country N Results Statistically significant differences only are reported Pang, Heffelfinger et al. 1989-1996 USA Planned home birth: Compared with women planning hospital births, women 2002 6,133 planning home births had more: Level of evidence III-3 Planned hospital low 5-minute Apgar scores≤ 3 (0.41% vs 0.20%) birth:10,593 neonates requiring assisted ventilation > 30 minutes (0.3% vs 0.2%). Ackermann-Liebrich, 1989-1991 Switzerland Planned home birth: Compared with planned hospital births, infants of mothers Voegli et al. 1996 489 who planned home births had higher mean 5-minute Apgar Level of evidence III-2 Planned hospital birth: scores (9.3 vs 9.0). 385 Woodcock, Read et al. 1981-1987 Australia Planned home birth: Compared with planned hospital births, fewer infants of 1994 976 mothers who planned home births had: Level of evidence III-2 Planned hospital birth: birthweight <2,500g (OR=0.47, 95% CI 0.25-0.88), 2,928 5-minute Apgar score <8 (OR=0.44, 95% CI 0.25-0.79), oxygen alone (OR=0.32, 95% CI 0.25-0.42), other resuscitation methods (OR=0.16, 95% CI 0.09-0.27), time to spontaneous respiration >1 minute (OR=0.45, 95% CI 0.32-0.63), birth trauma (OR=0.28, 95% CI 0.18-0.44), and were more likely to be post-term ≥42 weeks (OR=2.46, 95% CI 1.91-3.15). * birth trauma=subdural or cerebral hemorrhage, fracture of clavicle, long bones or skull, facial nerve injury, Erb palsy, or unspecified. ** resuscitation at birth = intermittent positive airway pressure via endotracheal tube or chest compression, or use drugs for resuscitation. *** IPPV = intermittent positive pressure ventilation. 32

Table 9. Perinatal mortality. Study Years of study Country N Results Statistically significant differences only are reported Hutton, Reitsma et al. 2003-2006 Ontario, Canada Planned home birth: No increased risk of perinatal and neonatal mortality among 2009 6,692 home births (1/1,000 births in both groups). Level of evidence III-2 Planned hospital birth: 6,692 de Jonge, van der Goes 2000-2006 Netherlands Planned home No increased risk of perinatal mortality among planned home et al. 2009 birth:321,303 compared with planned hospital birth. Level of evidence III-2 Planned hospital Intrapartum death: RR=0.97 (95% CI 0.69-1.37) birth:163,261 Intrapartum death and neonatal death ≤24hours: RR=1.02 (95% CI 0.77-1.36) Intrapartum death and neonatal death ≤7 days: RR=1.00 (95% CI 0.78-1.27) Kennare, Keirse et al. 1991-2006 South Australia Planned home birth: Compared with planned hospital birth, planned home birth had: 2009 1,141 increased risk of intrapartum death (1.8/1,000 vs 0.8/1,000) Level of evidence III-2 Planned hospital increased risk of death from intrapartum asphyxia birth: 297,192 (2.6/1,000 vs 0.3/1,000), No increase in risk of perinatal mortality: 7.9/1,000 vs 8.2/1,000 for home and hospital births. Janssen, Saxell et al. 2000-2004 British Columbia, Planned home birth: No increased risk of perinatal mortality among homebirths 2009 Canada 2,889 compared with midwife-led hospital birth and physician-led Level of evidence III-2 Planned midwife- hospital birth: 0.35/1,000 vs 0.57/1,000 births vs 0.64/1,000 led hospital birth: 4,752 Planned physician- led hospital birth: 5,331 Lindgren, Radestad et 1992-2004 Sweden Planned home birth: No increase in risk of intrapartum and neonatal mortality among al. 2008 897 home compared with hospital birth: 2.2/1,000 vs 0.7/1,000. Level of evidence III-2 Planned hospital birth:11,341 33

Study Years of study Country N Results Statistically significant differences only are reported Mori, Dougherty and 1994-2003 England and 4,991 intrapartum No difference in intrapartum perinatal mortality (IPPM) for Whittle 2008 Wales perinatal deaths booked home births compared with overall IPPM. Level of evidence IV from 6,314,315 Women booked for home birth but requiring transfer were at births highest risk. Overall IPPM 0.79/1,000 Booked home birth IPPM 1.28/1,000 Completed home birth IPPM 0.48/1,000 Unintended home birth IPPM 1.42/1,000 Transferred IPPM 6.05/1,000 Johnson and Daviss 2000 USA and Canada Planned home birth: No excess perinatal mortality among home births: 2.0/1,000. 2005 5,418 Level of evidence IV Janssen, Lee et al. 1998-1999 Canada Planned home birth: No excess perinatal mortality home birth compared with doctor- 2002 862 led hospital birth vs midwife-led hospital birth: Level of evidence III-2 Planned doctor-led 3.4/1,000 vs 1.3/1,000 vs 0 births. birth: 743 Planned midwife- led hospital birth: 571 Wolleswinkel-van den 1996-1997 Netherlands 319 deaths in all No evidence that sub-standard factors related to place of birth Bosch, Vredevoogd et birth settings caused differences in perinatal mortality. al. 2002 Level of evidence III-3 Pang, Heffelfinger et al. 1989-1996 USA Planned home birth: Increased risk of neonatal death among planned home 2002 6,133 compared with planned hospital births: Level of evidence III-3 Planned hospital RR 1.99 (95%CI 1.06-3.73). birth:10,593 Neonatal mortality 3.5/1,000 vs 1.7/1,000 live births. De Reu, Hijhuis et al. 1994-1995 Netherlands 73 deaths from all Perinatal mortality was not higher in home births. 2000 places of birth Level of evidence IV 34

Study Years of study Country N Results Statistically significant differences only are reported Murphy and Fullerton 1994-1995 USA Planned home birth: 2.5/1,000 1998 1,404 Level of evidence IV Bastian, Keirse et al. 1985-1990 Australia 50 deaths from 7.1/1,000 compared with all Australian births 10.8/1,000 1998 home births The majority of deaths occurred in women with obstetric risk Level of evidence IV factors. Northern Region 1981-1984 UK 14 deaths from Perinatal mortality rate was lower in home births compared with Perinatal Mortality planned home all births in the region 4.8/1000 vs 9.7/1000. Survey Coordinating births Group 1996 Level of evidence IV Woodcock, Read et al. 1981-1987 Western Australia Planned home birth: No increased risk of perinatal mortality among planned home 1994 976 home birth compared with hospital birth: 5.1/1,000 vs 4.1/1,000. Level of evidence III-2 Planned hospital birth: 2,928 Woodcock, Read et al. 1981-1987 Western Australia Planned home birth: Compared with all singleton Caucasian WA births, planned 1990 995 homebirths had: Level of evidence IV stillbirth rate of 2.0/1,000 vs 4.8/1,000 neonatal death rate of 3.0/1,000 vs 2.9/1,000 no difference in mortality. 35

References

Ackermann-Liebrich, U., Voegli, T., Gunter-Witt, K., Kunz, I., Zullig, M., Schindler, C., et al. (1996). Home versus hospital deliveries: follow up study of matched pairs for procedures and outcome. BMJ 313: 1313-1318.

Amelink-Verburg, M. P. and Buitendijk, S. E. (2010). Pregnancy and Labour in the Dutch Maternity Care System: What Is Normal? The Role Division Between Midwives and Obstetricians. J Midwifery Womens Health 55: 216-225.

Amelink-Verburg, M. P., Verloove-Vanhorick, S. P., Hakkenberg, R. M. A., Veldhuijzen, I. M. E., Bennebroek Gravenhorst, J. and Buitendijk, S. E. (2007). Evaluation of 280 000 cases in Dutch midwifery practices: a descriptive study. BJOG 115: 570-578.

Anthony, S., Buitendijk, S. E., Offerhaus, P. M., Dommelen, P. and Pal-de Bruin, K. M. (2005). Maternal factors and the probability of a planned home birth. BJOG 1126: 748-753.

Bastian, H., Keirse, M.J.N.C. and Lancaster, P.A.L. (1998). Perinatal death associated with planned home birth in Australia: population based study. BMJ 317: 384-388.

Christiaens, W., Bracke, P. (2009). Place of birth and satisfaction with childbirth in Belgium and the Netherlands. Midwifery 252: e11-19.

Crotty, M., Ramsay, A.T., Smart, R. and Chan, A. (1990). Planned homebirths in South Australia 1976-1987. MJA 153: 664-671.

Davis-Floyd, R. (2003). Home-birth emergencies in the US and Mexico: the trouble with transport. Soc Sci Med 56: 1911-1931.

de Jonge, A., van der Goes, B. Y., Ravelli, A. C., Amelink-Verburg, M. P., Mol, B. W., Nijhuis, J. G., et al. (2009). Perinatal mortality and morbidity in a nationwide cohort of 529,688 low-risk planned home and hospital births. BJOG 1169: 1177-1184.

De Reu, P. A. O. M., Hijhuis, J.G., Oosterbaan, H.P. and Eskes, T.K.A.B. (2000). Perinatal audit on avoidable mortality in a Dutch rural region: a retrospective study. Eur J Obst Gynecol Reprod Biol 88: 65-69.

Declercq, E. R., Paine, L.L. and Winter, M.R. (1995). Home birth in the United States, 1989-1992. J Nurse Midwifery 406: 474-482.

Department of Health. (2007). Improving Maternity Services: Working Together Across Western Australia. Department of Health, Perth, Western Australia.

Department of Health (2007) 12th Report of the Perinatal and Infant Mortality Committee of Western Australia, Deaths 2002-2004. Department of Health, Perth, Western Australia. 51

Department of Health. (2008). Rewiev of Homebirth in Western Australia. Department of Health, Perth, Western Australia.

Dowswell, T., Thornton, J.G.,Hewison, J and Lilford, R.J.L. (1996). Should there be a trial of home versus hospital delivery in the United Kingdom? BMJ 312: 753- 757.

Evers, A. C. C., Brouwers, H. A. A., Hukkelhoven, C. W. P. M., Nikkels, P. G. J., Boon, J., van Egmond-Linden, A., et al. (2010). Perinatal mortality and severe morbidity in low and high risk terms pregnancies in the Netherlands: prospective cohort study. BMJ 341: c5639.

Gee, V. (2010). Perinatal, Infant and Maternal Mortality in Western Australia, 2004- 2008, Department of Health, Perth, Western Australia.

Gyte, G., Dodwell, M., Newburn, M., Sandall, J., Macfarlane, A. and Bewley, S. (2009). Estimating intrapartum-related perinatal mortality rates for booked home births: when the 'best' available data are not good enough. BJOG 116: 933-942.

Gyte, G., Dodwell, M., Newburn, M., Sandall, J., Macfarlane, A. and Bewley, S. (2010). Safety of planned home births: Findings of meta-analysis cannot be relied on. BMJ 341: c4033.

Henderson, J., Hornbuckle, J. and Doherty, D. (2007). Models of Maternity Care: A Review of the Evidence. WA Department of Health. Perth.

Hendrix, M., Van Horck, M., Moreta, D., Nieman, F., Nieuwenhuijze, M., Severens, J., et al. (2009). Why women do not accept randomisation for place of birth: feasibility of a RCT in the Netherlands. BJOG 116: 537-542.

Hildingsson, I. M., Lindgren, H. E., Haglund, B., x00E, destad, I. J., Hildingsson, I. M., et al. (2006). Characteristics of women giving birth at home in Sweden: a national register study. AJOG 195: 1366-1372.

Howe, K. A. (1988). Home births in south-west Australia. MJA 149: 296-298.

Hutton, E. K., Reitsma, A. H., Kaufman, K., Hutton, E. K., Reitsma, A. H. and Kaufman, K. (2009). Outcomes associated with planned home and planned hospital births in low-risk women attended by midwives in Ontario, Canada, 2003-2006: a retrospective cohort study. Birth 363: 180-189.

Janssen, P. A., Lee, S.K., Ryan, E.M., Etches, D.J., Farquharson, D.F., Peacock, D., et al. (2002). Outcomes of planned home births versus planned hospital births after regulation of midwifery in British Columbia. CMAJ 166: 315-323.

Janssen, P. A., Saxell, L., Page, L. A., Klein, M. C., Liston, R. M., Lee, S. K., et al. (2009). Outcomes of planned home birth with registered midwife versus planned hospital birth with midwife or physician.[Erratum in CMAJ 181:617]. CMAJ 181: 377-383. 52

Johnson, K. C., and Daviss, B.A. (2005). Outcomes of planned home births with certified professional midwives: large prospective study in North America. BMJ 330: 1416-1422.

Kennare, R. M., Keirse, M. J., Tucker, G. R., Chan, A. C., Kennare, R. M., Keirse, M. J. N. C., et al. (2009). Planned home and hospital births in South Australia, 1991-2006: differences in outcomes. MJA 192: 76-80.

Keirse, M. J. C. (2010). Home birth: Gone Away, Gone Astray, and Here To Stay. Birth 37: 341-346.

Koehler N., U., Solomon D.,A., Murphy M. (1984). Outcomes of a rural Sonoma county home birth practice: 1976-1982. Birth 11: 165-169.

Lancet (2010). Editorial: Home birth - proceed with caution. Lancet 376: 303.

Lindgren, H. E., Radestad, I.J., Christensson, K., and Hildingsson, I.M. (2008). Outcome of planned home births compared to hospital births in Sweden between 1992 and 2004. A population-based register study. Acta Obstet Gynecol Scand 87: 751-759.

MacDorman, M. F., Declercq, E., Menacker, F. (2010). Trends and Characteristics of Home Births in the United States by Race and Ethnicity, 1990-2006. Birth 38: 1-7.

McMurtrie, J., Catling-Paul, C., Teate, A., Caplice, S., Chapman, M. and Homer, C. (2009). The St. George Homebirth Program: An evaluation of the first 100 booked women. ANZJOG 49: 631-636.

Mori, R., Dougherty, M. and Whittle, M. (2008). An estimation of intrapartum-related perinatal mortality rates for booked home births in England and Wales between 1994 and 2003. BJOG 1155: 554-559.

Murphy, P. A. and Fullerton, J. T. (1998). Outcomes of intended home births in nurse-midwifery practice: a prospective descriptive study. Obstet Gynecol 92: 461-470.

NHMRC. (2007). "Additional levels of evidence and grades for recommendations for developers of guidelines. Pilot program 2005-2007.", accessed on 2 Feb 2011 http://www.nhmrc.gov.au/_files_nhmrc/file/guidelines/levels_grades05.pdf.

Northern Region Perinatal Mortality Survey Coordinating Group (1996). Collaborative survey of perinatal loss in planned and unplanned home births. BMJ 313: 1306-1309.

Pang, J. W. Y., Heffelfinger, J.D., Huang, G.J., Benedetti, T.J. and Weiss, N.S. (2002). Outcomes of planned home births in Washington State: 1989-1996. Obstet Gynecol 100: 253-259.

Ravelli, A., Jager, K., de Groot, M., Erwich, J., Rijninks-van Driel, G., Tromp, M., Eskes, M., Abu-Hanna A. and Mol, B. (2010). Travel time from home to 53

hospital and adverse perinatal outcomes in women at term in the Netherlands. BJOG [Dec 8 Epub ahead of print].

Shearer J., M., L. (1985). Five year prospective survey of risk of booking for a home birth in Essex. BMJ 219: 1478-80.

Symon, A., Winter, C., Inkster, M., Donnan, P.T. (2009). Outcomes for births booked under an independent midwife and births in NHS maternity units: matched comparison study. BMJ 338: b2060.

Vedam, S. (2003). Home Birth versus Hospital Birth: Questioning the Quality of the Evidence on Safety. Birth 301: 57-63.

Wax, J. R., Lucas, F.L., Lamond, M., Pinette, M.G. (2010). Maternal and newborn outcomes in planned home birth vs planned hospital births: a metaanalysis. AJOG 203: 243.e1-8.

Wax, J. R., Pinette, M. G., Cartin, A., Blackstone, J., Wax, J. R., Pinette, M. G., et al. (2010). Maternal and newborn morbidity by birth facility among selected United States 2006 low-risk births. AJOG 202: 152.e1-5.

Western Australian Department of Health (2007). 12th Report of the Perinatal and Infant Mortality Committee of Western Australia, Deaths 2002-2004.

Western Australian Department of Health (2008). Review of homebirths in Western Australia.

Wiegers, T. A. (2009). The quality of maternity care services as experienced by women in the Netherlands. BMC Pregnancy & Childbirth 9: 9-18.

Wiegers, T. A., Keirse, M.J.N.C., van der Zee, J. and Berghs, G.A.H. (1996). Outcome of planned home and planned hospital births in low risk pregnancies: prospective study in midwifery practices in the Netherlands. BMJ 313: 1309- 1313.

Wiegers, T. A., van der Zee, J. and Keirse, M.J.N.C. (1998). Transfer from home to hospital: what is its effect on the experience of childbirth? Birth 251: 19-24.

Wolleswinkel-van den Bosch, J. H., Vredevoogd, C.B., Borkent-Polet, M., van Eyck, J., Fetter, W.P.F., Lagro-Janssen, T.L.M., et al. (2002). Substandard factors in perinatal care in The Netherlands. Acta Obstet Gynecol Scand 81: 17-24.

Woodcock, H. C., Read, A.W., Bower, C., Stanley, F.J. and Moore, D.J. (1994). A matched cohort study of planned home and hospital births in Western Australia 1981-1987. Midwifery 10: 125-135.

Woodcock, H. C., Read, A.W., Moore, D.J., Stanley, F.J. and Bower, C. (1990). Planned homebirths in Western Australia 1981-1987. MJA 153: 672-678.

Health Networks Branch Level 2C, 189 Royal St East Perth Western Australia 6004 56